INDEX

Explanations

words related to diverse experiences or concepts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "verse" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

urated

-0.84

rera

-0.80

arij

-0.78

ENDED

-0.78

OWER

-0.73

aul

-0.71

©¶æ¥µ

-0.70

URA

-0.69

govtrack

-0.68

chen

-0.68

POSITIVE LOGITS

verse

0.89

terday

0.86

backer

0.81

theless

0.78

lihood

0.73

 Sins

0.69

âĸ¬

0.68

xual

0.68

=-=-=-=-=-=-=-=-

0.64

 Ange

0.63

Activations Density 0.026%

words related to diverse experiences or concepts

instances of the word "verse" in various contexts

No Comments

No Known Activations